DESCRIPTION QUALITY TESTING METHOD AND PRODUCTION METHOD OF RARE EARTH MAGNET ALLOY INGOT, THE RARE EARTH MAGNET ALLOY INGOT, AND RARE EARTH MAGNET ALLOY TECHNICAL FIELD The present invention relates to a method for determining quality of an ingot of a rare earth magnet alloy (referred to as an RE-TM-B magnet alloy) having the following composition: RE (RE is at least one metallic element selected from among lanthanoids, including Y (i. e., Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu) in an amount of 27-34 mass%; B (boron) in an amount of 0.7-1.4 mass%; and TM as the balance (TM represents a metal comprising a transition metal, including Fe, as an essential element). The invention also relates to a method for producing the ingot, to a rare earth magnet alloy ingot, and to a rare earth magnet.

BACKGROUND ART Recently, there has been increasing demand for RE-TM-B magnet alloys for use in voice coil motors (VCMs) employed in a hard disk drive of a personal computer, magnetic resonance imaging (MRI) devices for medical application, and motors. In association with the trend of enhancing performance and down-scaling of these devices, further enhancement of magnets themselves is desired.

Sintered magnets employing an RE-TM-B magnet alloy are produced according to the following procedure. Specifically, a rare earth metal or a matrix alloy such as REFe (RE-Fe binary alloy) serving as an RE source; solid boron or ferroboron serving as a boron (B) source; pure iron or Atomiron serving as a TM source; and other additive elements are appropriately used. These sources are melted in an alumina crucible placed in vacuum or an inert gas atmosphere, and the resultant molten alloy is cast, to thereby yield an alloy ingot. The thus-yielded alloy ingot is pulverized under nitrogen or in an inert gas atmosphere, to thereby form a powder of the alloy having a particle size of approximately some pjn. During or after pulverization, a liquid or solid (powder form) lubricant is added to the alloy. The thus-obtained alloy powder is molded in a magnetic field, and the compact is sintered in vacuum or in an inert gas atmosphere, to thereby produce a sintered compact. The shape of the resultant sintered compact is adjusted, and the surface of the compact is plated with nickel or aluminum so as to prevent corrosion or erosion, to thereby obtain a sintered magnet as a final product.

Among magnetic characteristics of sintered magnets, residual magnetization, coercive force, and a squareness property are of particular importance. For enhancing residual magnetization, crystal orientation of each particle of the powder and density of the sintered compact are preferably higher. For enhancing coercive force, the particle size of the pulverized alloy must be reduced. For improving the squareness property, the particle size distribution profile of the powder must be sharp. In other words, magnetic characteristics of a sintered magnet greatly depend on properties of the alloy powder for producing the magnet.

The alloy powder is produced according to the following procedure. Specifically, the cast alloy ingot is coarsely crushed into fragments preferably having a size of approximately some cm (in the present specification, the term"coarsely crushing"refers to fragmentation of the alloy ingot into fragments having a size of 0.1-10 cm). The fragments are placed in an airtight container, and the interior of the container is adjusted to vacuum. Subsequently, hydrogen is introduced into the container, to thereby maintain the fragments in a hydrogen atmosphere, and cracking is induced to break the alloy fragments based on the phenomenon that the alloy expands when it absorbs hydrogen.

This operation is called hydrogen decrepitation.

Hydrogenation occurs more readily in the phase rich in RE (hereinafter referred to as RE-rich phase) dispersed in the main phase of an RE-TM-B magnet alloy than in the main phase. Hydrogen decrepitation, employing the above property, is a step in which cracks are generated through expansion accompanying hydrogenation, and cracking propagates from the alloy ingot surface in a chain-transfer manner, to thereby fragment the ingot.

Subsequently, the alloy ingot which has been subjected to hydrogen decrepitation is further coarsely crushed by use of a crusher such as a Brawn mill, to thereby form a powder having a size of some hundreds of m, followed by micro-pulverization by use of a crusher such as a jet mill, to thereby reduce the size to approximately some urn.

Characteristics required of the alloy powder are considered to preferably meet all the following conditions.

1) A first condition is that a single particle of the powder does not contain a plurality of crystals. The condition is important for arranging crystal orientation in one direction during application of a magnetic field to the powder. If one particle has a plurality of crystals having different crystal orientations, crystal orientation of the particle as a whole is arranged in the direction corresponding to the sum of crystal axis vectors in the particle during application of a magnetic field, failing to attain high crystal orientation.

2) A second condition is that RE-rich phase is present on the surface of each particle of the powder, and that the powder contains no particles which are formed exclusively of RE-rich phase. The condition is considerably important, in that the RE- rich phase plays an important role as a liquid phase during liquid phase sintering. In other words, in order to produce a high-density sintered product of the alloy powder by carrying out homogeneous liquid phase sintering, the liquid phase is preferably distributed homogeneously in the molded compact. If RE-rich phase is caused to be present exclusively on the surface of each particle of the powder, the liquid phase can be distributed almost homogeneously. In the case in which RE-rich phase is present inside the particles, portions of RE-rich phase which are not involved in liquid phase sintering are generated, thereby failing to attain effective utilization of RE-rich phase. When some particles are formed exclusively of RE-rich phase, the distribution profile of RE-rich phase becomes broad and dispersion of RE-rich phase becomes poor, thereby failing to attain highly homogeneous RE-rich phase distribution.

3) A third condition is that the powder has a particle size, as measured by use of a Fisher Sub-Sieve Sizer, of approximately 3-4 um and a narrow particle size distribution.

The characteristics of the sintered compact obtained by molding the powder and sintering are varied in accordance with the particle size of the powder. When the particle size distribution is broad, micro-particles included in the powder increases the activity of the powder, to thereby disadvantageously elevate oxygen concentration of the produced magnet, whereas when the powder contains particles of large size or the particle size is 5 pm or more, magnetic characteristics of the produced magnet, particularly coercive force, are deteriorated.

By performing hydrogen decrepitation prior to mechanical pulverization, micro- cracks can be generated in advance in the alloy ingot along RE-rich phase present in the grain boundaries and particles. The particle size of the produced powder is determined in accordance with the metallographic structure of the alloy. Thus, a rare earth magnet alloy ingot having an appropriate alloy metallographic structure is subjected to hydrogen decrepitation and then pulverized, to thereby obtain an alloy powder satisfying all the aforementioned conditions 1), 2), and 3).

One preferred method for casting an alloy suitable for producing an alloy powder having a preferable particle size distribution is a strip casting method (hereinafter referred to as an SC method). In the SC method, a molten alloy is poured onto a copper roll, to thereby cast the alloy into strips. The thus-cast alloy strips are introduced into a container for collecting the strips, and the cooling rate thereof is controlled. As disclosed in Japanese Unexamined Patent Application, First Publication, No. 09-170055, upon cooling, the cooling rate of the alloy strips is preferably controlled to 300°C/sec or more within a range of the melting temperature to 800°C and to 10°C/sec or less within a range of 800-600°C.

In contrast, when an alloy ingot produced through a conventional book mold method is pulverized, production of powder particles formed exclusively of RE-rich phase is very likely, thereby failing to obtain favorable particles.

Even when the SC method is employed, deviation of the aforementioned conditions in connection with the cooling rate is not preferred. The reason for this is as follows.

When the cooling rate within a range of 800-600°C is in excess of 10°C/sec, RE- rich phase is distributed more minutely. By subjecting such SC pieces to hydrogen decrepitation, expansion due to hydrogenation of RE-rich phase is reduced. Thus, the rate of crack generation in the SC pieces decreases. Accordingly, the following problems arise. a) As compared with SC pieces which have been cast under preferred conditions, the above SC pieces require a longer time for hydrogen decrepitation. If the time is short, portions in which no cracks are generated remain in the pieces. Thus, the powder produced by pulverizing the pieces tends to contain RE-rich phase present not on the surface of particles but inside the particles. b) Even if hydrogen decrepitation is performed for a sufficient period of time to thereby generate cracks, cracking along the RE-rich phase is generated excessively minutely, to thereby produce a powder having an excessively reduced particle size. Thus, such an alloy powder is readily oxidized, and flowability of the powder is highly prone to decrease considerably.

In contrast, when the cooling rate within a range of 800-600°C is 0.5°C/sec or less, the RE-rich phase is more sparsely dispersed, and where the RE-rich phase is found, it tends to present a dense phase. Accordingly, the following problems arise. c) Although hydrogen decrepitation of RE-rich phase can be completed within a very short period of time, cracks generated by hydrogen decrepitation are dispersed considerably sparsely, thereby producing a powder having a larger particle size. Even when the particle size has been successfully adjusted through mechanical pulverization, the degree of uniformity in coverage of the surface of the particles with RE-rich phase decreases, and particles formed exclusively of RE-rich phase are incorporated in the powder at higher probability.

Thus, depending on the metallographic structure of the rare earth magnet alloy ingot, the particle size distribution of the alloy powder which has been subjected to hydrogen decrepitation varies, and hydrogen absorption behavior of the alloy also changes.

As described above, hydrogen decrepitation performed prior to mechanical pulverization is an important issue in production of an alloy powder having a particle size distribution suitable for forming a sintered magnet having excellent magnetic characteristics from a cast ingot of an RE-TM-B magnet alloy.

However, conventionally, no definite method has been devised for determining quality of a rare earth magnet alloy ingot so as to quantitatively evaluate the degree of hydrogen decrepitation of an RE-TM-B magnet alloy ingot and correlate the evaluation results with magnetic characteristics.

The relationship between the metallographic structure of an alloy and cooling behavior has already been reported. For example, Japanese Unexamined Patent Application, First Publication, No. 08-269643 discloses the relationship between the metallographic structure of an alloy and primary and secondary cooling rates, and Japanese Unexamined Patent Application, First Publication, No. 09-170055 discloses the relationship between the metallographic structure of an alloy and cooling behavior within a temperature range of 800-600°C. However, these references fail to mention how variation in hydrogen decrepitation behavior affects the characteristics of the alloy powder and, further, they fail to mention how variation in hydrogen decrepitation behavior affects magnetic characteristics of the magnet produced from the powder.

During hydrogen decrepitation of a rare earth magnet alloy ingot, fragmentation behavior is controlled by RE-rich phase present in the alloy, and therefore, the distribution profile of the RE-rich phase is very important. However, the particle size distribution of the alloy powder after completion of hydrogen decrepitation and magnetic characteristics of the sintered magnet produced in a subsequent step are very difficult to predict by a conventional method for evaluating the distribution profile of RE-rich phase in a rare earth magnet alloy ingot. Thus, disadvantageously, quality of a cast alloy ingot of a rare earth magnet alloy cannot be determined until a magnet is actually produced from the ingot.

In contrast, the present inventor has found that, if a cast alloy ingot of a rare earth alloy exhibiting suitable hydrogen absorption behavior can be produced, an alloy powder having a preferred particle size distribution can be produced through hydrogen decrepitation, and a sintered magnet of excellent magnetic characteristics can be produced.

DISCLOSURE OF INVENTION Thus, an object of the present invention is to provide a method for determining quality of a rare earth magnet alloy ingot, for selection of an RE-TM-B magnet alloy ingot which can provide an alloy powder producing a sintered magnet of excellent magnetic characteristics. Another object of the invention is to provide a method for producing a rare earth magnet alloy ingot including the method for determining quality. Still another object is to provide a rare earth magnet alloy ingot for producing a sintered magnet of excellent magnetic characteristics. Furthermore, another object is to provide a rare earth magnet fabricated by use of the powder produced by pulverizing the alloy ingot.

Accordingly, a first aspect of the present invention is to provide a method for determining quality of a rare earth magnet alloy ingot having a composition of RE (RE is at least one metallic element selected from lanthanoids, including Y (i. e., Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu) in an amount of 27 to 34 mass%; B (boron) in an amount of 0.7 to 1.4 mass%; and TM as a balance (TM represents a metal comprising a transition metal, including Fe, as an essential element), the method comprising steps of holding the ingot in a reduced-pressure atmosphere, subsequently placing the ingot in a hydrogen atmosphere, and determining a hydrogen absorption behavior of the ingot while the ingot is held in the hydrogen atmosphere.

In addition, the hydrogen absorption behavior of the rare earth magnet alloy ingot may be determined by measuring a time-elapsed change in an amount of hydrogen absorbed in the ingot from a time the ingot is place in the hydrogen atmosphere.

In addition, the rare earth magnet alloy ingot may be coarsely crushed and is then held in a reduced-pressure atmosphere In addition, the rare earth magnet alloy ingot may be held in a reduced-pressure atmosphere at a pressure of 8 x 10-4 to 1 x 1 o-2 Pa.

In addition, the rare earth magnet alloy ingot may be placed in a hydrogen atmosphere at a temperature of 273 to 373 K.

In addition, the rare earth magnet alloy ingot may be placed in a hydrogen atmosphere at a pressure of 101 to 160 kPa.

In addition, the rare earth magnet alloy ingot may be produced by a rapid-cool casting method.

In addition, the rapid-cool casting method may be a strip casting method.

Furthermore, the hydrogen absorption behavior of the rare earth magnet alloy ingot may be determined by measuring a period of time between an instant the rare earth magnet alloy ingot is placed in the hydrogen atmosphere and a point in time when an amount of absorbed hydrogen reaches I % of a maximum amount of hydrogen absorbable in the rare earth magnet ingot.

Furthermore, a second aspect of the present invention is to provide a method for producing a rare earth magnet alloy ingot comprising steps of determining quality of a rare earth magnet alloy ingot by employing a method for determining quality of a rare earth magnet alloy ingot according to the first aspect of the present invention, and removing a rare earth magnet alloy ingot having unsatisfactory quality at a step of magnet production.

Furthermore, a third aspect of the present invention is to provide a rare earth magnet alloy ingot having a composition of RE (RE is at least one metallic element selected from lanthanoids including Y (i. e., Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu) in an amount of 27 to 34 mass%; B (boron) in an amount of 0.7 to 1. 4 mass%; and TM as a balance (TM represents a metal comprising a transition metal, including Fe, as an essential element), wherein, when the rare earth magnet alloy ingot is maintained in a reduced-pressure atmosphere of 8 x 104 to 1 x 10-2 Pa and the ingot is subsequently placed in a hydrogen atmosphere at a pressure of 101 to 160 kPa and maintained at 283 to 313 K, a hydrogen absorption behavior of the ingot is determined by a period of time between a time the ingot is placed in the hydrogen atmosphere and a point in time when an amount of absorbed hydrogen reaches 1% the maximum absorbable amount of hydrogen in the alloy of 200 to 2,400 seconds and a maximum hydrogen- absorption rate of the alloy of 1.0 x 10-4 to 1.2 x 10-3 mass%/sec In addition, the rare earth magnet alloy ingot may be coarsely crushed and is then held in a reduced-pressure atmosphere.

In addition, the rare earth magnet alloy ingot may be produced by a rapid-cool casting method.

In addition, the rapid-cool casting method may be a strip casting method.

Furthermore, a fourth aspect of the present invention is to provide a rare earth magnet produced from a rare earth magnet alloy ingot according to the third aspect of the present invention.

Furthermore, a fifth aspect of the present invention is to provided a rare earth magnet alloy ingot having a composition of RE (RE is at least one metallic element selected from among lanthanoids, including Y (i. e., Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu) in an amount of 27 to 34 mass% (a total amount of Eu, Gd, Tb, Dy, Ho, Yb, and Lu is limited to less than 1 mass%); B (boron) in an amount of 0.7 to 1.4 mass%; and TM as the balance (TM represents a metal comprising a transition metal, including Fe, as an essential element), wherein, when the rare earth magnet alloy ingot is held in a reduced-pressure atmosphere of 8 x 10-4 to 1 x 10'2 Pa and the ingot is subsequently placed in a hydrogen atmosphere at a pressure of 101 to 160 kPa and is held at 283 to 313 K, the hydrogen absorption behavior is determined by a period of time between a time the ingot is placed in the hydrogen atmosphere and a point of time when an amount of absorbed hydrogen reaches 1 % of a maximum amount of hydrogen absorbable in the alloy of 100 to 1,800 seconds and a maximum hydrogen-absorption rate of the alloy of 1.2 x 104 to 1.5 x10-3 mass%/sec.

In addition, the rare earth magnet alloy ingot may be coarsely crushed and is then held in a reduced-pressure atmosphere.

In addition, the rare earth magnet alloy ingot may be produced by a rapid-cool casting method.

In addition, the rapid-cool casting method may be a strip casting method.

Furthermore, a sixth aspect of the present invention is to provide a rare earth magnet produced from a rare earth magnet alloy ingot according to the fifth aspect of the present invention.

BRIEF DESCRIPTION OF DRAWINGS Fig. 1 is a schematic graph showing time-elapsing change in amount of hydrogen absorbed in a rare earth alloy ingot.

Fig. 2 is a schematic graph showing time-elapsing change in hydrogen absorption rate of a rare earth alloy ingot.

Fig. 3 is a graph showing comparison of different types of time-elapsing changes in amount of hydrogen absorbed in a rare earth alloy ingot.

Fig. 4 is a graph showing comparison of different types of time-elapsing changes in hydrogen absorption rates in a rare earth alloy ingot.

Fig. SA is a graph showing the relationship between T and BHmax of alloy A.

Fig. 5B is a graph showing the relationship between rmax and BHmax of alloy A.

Fig. 6A is a graph showing the relationship between T and BHmax of alloy B.

Fig. 6B is a graph showing the relationship between rmax and BHmax of alloy B.

Fig. 7A is a graph showing the relationship between T and BHmax of alloy C.

Fig. 7B is a graph showing the relationship between rmax and BHmax of alloy C.

Fig. 8A is a graph showing the relationship between T and BHmax of alloy D.

Fig. 8B is a graph showing the relationship between rmax and BHmax of alloy D.

Fig. 9A is a graph showing the relationship between T and BHmax of alloy E.

Fig. 9B is a graph showing the relationship between rmax and BHmax of alloy E.

Fig. I OA is a graph showing the relationship between T and BHmax of alloy F.

Fig. 1 OB is a graph showing the relationship between rmax and BHmax of alloy F.

BEST MODE FOR CARRYING OUT THE INVENTION The present inventor has investigated the hydrogen absorption behavior of RE- TM-B magnet alloy ingots, and has identified the characteristics in terms of the hydrogen absorption behavior of a rare earth magnet alloy ingot suitable for producing an alloy powder from which a sintered magnet of excellent magnetic characteristics can be produced. Specifically, by determining the hydrogen absorption behavior under the following conditions, a determination can be made as to whether the RE-rich phase is suitably distributed in the rare earth magnet alloy ingot and the alloy ingot can be suitably cracked along the RE-rich phase through hydrogen decrepitation.

Specifically, a rare earth magnet alloy having the following composition: RE (RE is at least one metallic element selected from among lanthanoids, including Y (i. e., Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu) in an amount of 27-34 mass%; B (boron) in an amount of 0.7-1.4 mass%; and TM as the balance (TM represents a metal comprising a transition metal, including Fe, as an essential element) and produced preferably through a rapid-cool casting method is maintained in an airtight container in a reduced-pressure atmosphere, preferably at 8 x 10-4-1 x l0-2 Pa; and subsequently, the ingot is placed in a hydrogen atmosphere preferably maintained at 273-373 K and 101-160 kPa. By investigating the variation in hydrogen absorption behavior of the alloy ingot during maintenance thereof, the quality of the rare earth magnet alloy ingot can be determined.

According to the aforementioned method for determining quality of a rare earth magnet alloy ingot, there have been identified characteristics of a rare earth magnet alloy ingot capable of producing a sintered magnet of excellent magnetic characteristics.

Specifically, the hydrogen absorption behavior of a rare earth magnet alloy ingot capable of producing a sintered magnet of excellent magnetic characteristics is characterized by the following. That is, when the aforementioned rare earth magnet alloy ingot is maintained in a reduced-pressure atmosphere of 8 x 10'4-1 x 10-2 Pa and hydrogen (101- 160 kPa) is introduced into the atmosphere while the temperature is maintained at a predetermined temperature falling within 283-313 K, the alloy ingot exhibits a period of time between the instant the ingot is placed in the hydrogen atmosphere and the point of time when the amount of absorbed hydrogen reaches 1 % the maximum absorbable amount of hydrogen in the alloy of 200-2,400 seconds and a maximum hydrogen-absorption rate of the alloy falling within a range of 1.0 x 10-4-1. 2 x 10-3 mass%/sec.

With respect to rare earth magnet alloy ingots having the following composition: RE (RE is at least one metallic element selected from among lanthanoids, including Y (i. e., Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu) in an amount of 27-34 mass% (the total amount of Eu, Gd, Tb, Dy, Ho, Yb, and Lu is limited to less than 1 mass%); B (boron) in an amount of 0.7-1.4 mass%; and TM as the balance (TM represents a metal comprising a transition metal, including Fe, as an essential element) and produced preferably through a rapid-cool casting method, the hydrogen absorption behavior of a rare earth magnet alloy ingot capable of producing a sintered magnet of excellent magnetic characteristics is characterized by the following. That is, when a rare earth magnet alloy ingot is maintained in a reduced-pressure atmosphere of 8 x 10-4-I x 10-2 Pa and hydrogen (101-160 kPa) is introduced into the atmosphere while the temperature is maintained at a predetermined temperature falling within 283-313 K, the alloy ingot exhibits a period of time between the instant the ingot is placed in the hydrogen atmosphere and the point of time when the amount of absorbed hydrogen reaches 1 % the maximum absorbable amount of hydrogen in the alloy of 100-1,800 seconds and a maximum hydrogen-absorption rate of the alloy falling within a range of 1. 2 x 104-1. 5 x 10-3 mass%/sec.

Thus, in the method of the present invention for producing a rare earth magnet alloy ingot, the quality of a rare earth magnet alloy to be produced is predicted by use of the alloy ingot according to the method of the present invention for determining quality of a rare earth magnet alloy ingot, and alloy ingots of unsatisfactory quality are removed from a magnet production step, to thereby enhance efficiency of production of excellent rare earth magnet alloy ingots.

In a preferred mode of the aforementioned method for producing a rare earth magnet alloy ingot, an RE-TM-B magnet alloy is cast through a rapid-cool casting method, to thereby provide an alloy metallographic structure which satisfies the aforementioned hydrogen absorption conditions.

Examples of applicable rapid-cool casting methods include a gas-atomization method, a spray-forming method, and a strip casting method, with a strip casting method (hereinafter referred to as an SC method) being particularly preferred.

Upon rapid-cool casting, among casting conditions, the average cooling rate is controlled to 300°C/sec or more within a range of the temperature of the molten alloy (e. g., 1,400°C) to 1,000°C and to 0.5-10°C/sec within a range of 800-600°C. More preferably, the average cooling rate is controlled to 500°C/sec or more within a range of the temperature of the molten alloy to 1,000°C and to 0.5-5.0°C/sec within a range of 800- 600°C.

The average cooling rate upon rapid-cool casting is measured in accordance with the following procedure. For example, when an SC method is employed, the main phase of the alloy is solidified on a roll which is rotating. Thus, the temperature of the molten alloy immediately before it falls onto the roll is measured by use of an immersed thermocouple, and the temperature of the molten alloy in which the main phase is being solidified while moving on the roll is measured by use of a dichroismic radiation pyrometer. The difference between two these temperatures is divided by the corresponding time, to thereby calculate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C (the temperature at which the main phase is completely solidified in the alloy). In a product collection section, the product has an initial temperature, which depends on the composition of the alloy, falling within 700- 900°C, and is gradually cooled. Thus, the time-elapsed change in temperature of the alloy residing in the production collection section is determined, to thereby derive a average cooling rate within 800-600°C.

A preferred mode the present invention for carrying out the casting method to produce a rare earth magnet alloy ingot will next be described.

Firstly, alloy raw materials are mixed to thereby obtain the following composition: RE = 27-34 mass%; B = 0.7-1.4 mass%; and TM = balance. The raw material mixture is heated in a vacuum chamber, in a vacuum atmosphere, or in an inert gas atmosphere, to thereby produce a molten alloy.

Taking an SC method as an example, a rapid-cool casting method will next be described.

The apparatus employed in the SC method includes a tundish for bringing a molten alloy into contact with a copper roll; a roll for rapid cooling for rapidly cooling the molten alloy; and a container for collecting the solidified alloy, and these members are accommodated in a vacuum chamber.

A molten alloy is poured into the tundish, and the molten alloy is poured from the tundish onto the roll for rapid cooling, to thereby cast the molten alloy. The cooling rate within a range of the temperature of the molten alloy (e. g., 1,400°C) to 1,000°C corresponds to the cooling rate of the molten alloy moving on the roll for rapid cooling.

The cooling rate of the molten alloy moving on the roll for rapid cooling can be adjusted by modifying the peripheral velocity of the roll for rapid cooling. For example, the greater the thickness of the molten alloy controlled by reducing the peripheral velocity of the roll for rapid cooling, the slower the cooling rate. The thus-solidified alloy is collected in the container, and the cooling rate of the alloy within 800-600°C is controlled by a predetermined method; e. g., maintaining temperature in the container or allowing an inert gas to flow through the container. Conventionally, in the case of casting of an alloy through a rapid-cooling method such as an SC method, the cooling rate only within the range of the temperature of the molten alloy to the solidification temperature of the alloy has been taken into consideration. In contrast, in the present invention, a rapidly cooled alloy which exhibits a suitable hydrogen absorption behavior can be obtained by controlling the cooling rate within 800-600°C.

A preferred alloy composition of a rare earth magnet alloy of the present invention will next be described.

With respect to rare earth elements, Sm, Er, and Tm are preferably not included.

These elements exhibit longitudinal anisotropy when an RE2TMI4B compound is formed, to thereby deteriorate magnetic anisotropy.

The amount of Al is difficult to control to 0.05 mass% or less, since Al unavoidably migrates into a molten alloy from a crucible employed in casting. Although Al effectively enhances coercive force, addition of an excessive amount of Al results in a decrease in residual magnetization. Thus, the amount is preferably controlled to 3 mass% or less.

Cu, having an effect of enhancing coercive force, is preferably added. However, addition of excessive Cu results in a decrease in residual magnetization. Thus, the amount is preferably controlled to 3 mass% or less.

The amount of oxygen is difficult to control to 0.02 mass% or less, since oxygen unavoidably migrates into raw material or a molten alloy during casting. An excessive amount of oxygen adversely affects magnetic characteristics. Thus, the amount is preferably controlled to 1 mass% or less.

The amount of carbon is difficult to control to 0.005 mass% or less, since carbon unavoidably migrates into raw material or a molten alloy during casting. A highly excessive amount of carbon adversely affects magnetic characteristics. Thus, the amount is preferably controlled to 0.2 mass% or less.

A preferred method of hydrogen decrepitation employed in the present invention for determining the hydrogen absorption behavior will next be described.

Preferably, the apparatus employed for carrying out hydrogen decrepitation for determining the hydrogen absorption behavior can maintain temperature; is adaptable to vacuum control by means of a sliding vane rotary vacuum pump or a hydraulic diffusion pump; and is durable to an internally applied pressure of approximately 200 kPa. The alloy sample to be processed is slightly fragmented, preferably to approximately 1-3 mm, so as to remove a certain amount of unfavorable oxide film covering the surface of the sample, to thereby develop virgin cut surfaces. In order to suppress variation in measuring temperature caused by heat accompanying hydrogen absorption, SC pieces are placed in a sample container such that the pieces are thinly spread, without being overlapped, or so that a maximum of one or two pieces are overlapped. The container is set in the apparatus in an airtight manner. The internal pressure is lowered to approximately 8 x 10-4 to 1 x 10-2 Pa, and the sample is maintained in the reduced- pressure atmosphere for a predetermined time (e. g., approximately three hours).

Hydrogen is introduced into the apparatus to a pressure of 101-160 kPa, preferably 101- 140 kPa, while the inside temperature of the apparatus is maintained at a predetermined temperature falling within 273-373 K, preferably 283-313 K. The point of time at which the sample starts being maintained in the hydrogen atmosphere is defined as the initial time, and the sequentially occurring time-elapsed change in pressure inside the apparatus is measured.

Although the temperature during the measurement varies in accordance with the circumstances, the temperature preferably falls within 283-313 K. When the temperature is 283 K or lower, particularly 273 K or lower, hydrogen absorption by an alloy occurs slowly, and hydrogen absorption requires a considerably long period of time, leading to deterioration in efficiency, whereas when the temperature is 313 K or higher, particularly 373 K or higher, hydrogen absorption reaction of the alloy proceeds at excessive speed, leading to difficulty in quality determination. In order to compare hydrogen absorption behavior of samples, hydrogen decrepitation must be carried out at the same temperature.

Among conditions of the reduced-pressure atmosphere, a reduced pressure of 1 x 10-2 Pa or higher is not sufficient for removing molecules of water and gases adhering to the surface of the alloy ingots, thereby retarding overall hydrogen absorption, whereas attaining a reduced pressure of 8 x 104 Pa or less requires a considerably long period of time even when a hydraulic diffusion pump is employed. Such a long time is not preferred, from a viewpoint of determination efficiency. Thus, the reduced-pressure atmosphere condition is preferably controlled to 8 x 104-1 x 10-2 Pa.

In hydrogen atmosphere conditions, when the pressure is 160 kPa or higher, hydrogen absorption reaction of the alloy ingots occurs at excessive speed, leading to difficulty in quality determination. When the pressure is 101 kPa or lower, a long period of time is required for determination, due to slow hydrogen absorption reaction, and the internal pressure of the apparatus becomes lower than the external pressure, leading to migration of air into the apparatus caused, for example, by trouble in the apparatus and in some cases forming detonating gas. This is also disadvantageous. Thus, the hydrogen pressure is preferably controlled to 101-160 kPa.

On the basis of the aforementioned time-elapsed change in pressure in the apparatus during hydrogen decrepitation, the time-elapsed change in amount of hydrogen absorbed in rare earth magnet alloy ingots (hydrogen absorption behavior) is calculated.

Data are plotted on a graph, to thereby obtain a curve. Fig. 1 shows a schematic view of the curve profile. From this curve, the maximum absorbable amount of hydrogen, corresponding to the amount of absorbed hydrogen which does not further increase due to saturation of absorption, is obtained. The period of time"T"between the start of hydrogen pressurization and the point of time when the amount of absorbed hydrogen reaches 1 % the maximum absorbable amount of hydrogen in the rare earth magnet alloy ingot is calculated. In addition, the gradient of each of the tangents with respect to the curve in Fig. 1 is calculated, and the time-elapsed changes in gradient are plotted in another graph, to thereby provide a graph showing the time-elapsed change in hydrogen- absorption rate of the alloy. Fig. 2 shows a schematic view of the profile. Since a curve of this type usually has a peak, the maximum value"rmax"of hydrogen-absorption rate can be calculated by reading the height of the peak. By employing the thus-obtained two indices, T and rmax, the condition of the rare earth alloy is evaluated, and the evaluation is used to make a determination as to whether or not magnetic characteristics suitable for sintered magnet can be attained.

In the present specification, the amount of hydrogen absorbed in a rare earth magnet alloy ingot is represented by the ratio (percentage) of the mass of hydrogen absorbed in the rare earth magnet alloy ingot to the mass of the ingot. Accordingly, the unit for the amount of absorbed hydrogen is mass%. In addition, the present inventor defines the absorbable amount of hydrogen experimentally as the amount of absorbed hydrogen which has reached saturation and does not change and at which the hydrogen- absorption rate has decreased to approximately 5 x 10-6 mass%/sec or less.

Figs. 3 and 4 show the results for some samples. Fig. 3 is a graph showing the time-elapsed change in amount of hydrogen absorbed in the alloy, and Fig. 4 is a graph showing the time-elapsed change in hydrogen-absorption rate of the alloy. In Figs. 3 and 4, each of reference numerals (1), (2), and (3) represents an alloy exhibiting a cooling rate during rapid-cool casting within 800-600°C, which differs from that of another alloy.

The order of the cooling rates is: (3) > (2) > (1). As is clear from these Figs., the lower the cooling rate within 800-600°C, the smaller"T"and the greater"rmax."This tendency has been identified for all relevant alloys, regardless of the alloy composition.

In addition, these alloys were pulverized, to thereby produce a magnet, and magnetic characteristics were investigated. Through investigation, differences in magnetic characteristics of the alloys have been identified, and deterioration in magnetic characteristics of an alloy which exhibits"T"and"rmax"falling outside the appropriate ranges has been identified.

When characteristics of the alloys are compared, the alloy composition, particularly the RE content, must be fixed. In addition, when the total amount of Eu, Tb, Dy, Ho, Er, Tm, Yb, and Lu is 1 mass% or more, the behavior of expansion of RE-rich phase caused by formation of hydrides changes, leading to prolongation of"T"and reduction of"rmax."Thus, it must be noted that ranges of"T"and"rmax"suitable for attaining excellent magnetic characteristics vary in accordance with the above total amount.

According to the present invention, the degree of hydrogen decrepitation of a rare earth magnet alloy ingot can be evaluated quantitatively, and magnetic characteristics of a sintered magnet to be produced can be predicted on the basis of the hydrogen absorption behavior during hydrogen decrepitation employed for determining quality of a rare earth magnet alloy ingot. Specifically, the alloy ingot is maintained in a reduced-pressure atmosphere; the ingot is subsequently placed in a hydrogen atmosphere; and the period of time"T"between the instant the ingot is placed in the hydrogen atmosphere and the point of time when the amount of absorbed hydrogen reaches 1 % the maximum absorbable amount of hydrogen in the rare earth magnet ingot is determined, along with the maximum hydrogen-absorption rate"rmax"of the rare earth magnet alloy ingot. The prediction is considered to be based on the criteria that the aforementioned time"T"and"rmax"vary in accordance with the distribution condition of the RE-rich phase in the rare earth magnet alloy ingot and can serve as indices for accurately predicting the distribution condition of the RE-rich phase.

Examples (Example 1) Alloy raw materials were provided and mixed to thereby obtain the following composition: Nd = 30.0 mass%; B = 0.98 mass%; Al = 0.3 mass%; Cu = 0.03 mass%; and iron = balance (hereinafter the alloy of this composition is referred to as alloy A). The raw material mixture was placed in an alumina crucible and heated in a vacuum chamber, initially in a vacuum atmosphere, then in an argon gas atmosphere, to thereby produce a molten alloy. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a copper roll for rapid cooling. The peripheral velocity of the roll was controlled to 1.2 m/sec. The average cooling rate of the alloy during casting within a range of the temperature of the molten alloy (approximately 1,400°C) to 1,000°C was calculated according to the following procedure. Specifically, the temperature of the molten alloy residing in the tundish was measured by use of an immersed thermocouple, and the temperature of the alloy having moved from a falling point on the roll to a position corresponding to a rotation of the roll over 60° was measured by use of a dichroismic radiation pyrometer. The difference between two measured temperatures was divided by the time of rotation of the roll over 60°, to thereby calculate the average cooling rate.

Through this procedure, the average cooling rate within a range of the temperature of the molten alloy to 1,000°C was found to be 800°C/sec. The thus-cast alloy was collected in a container for accommodating the alloy. The average cooling rate within 800-600°C was obtained by determining the time-elapsed change in temperature of the alloy residing in the container and dividing the determined change in temperature by the time required for the change from 800 to 600°C. The thus-obtained average cooling rate was 0.5°C/sec.

The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus for this treatment had an inner volume of 0.010 m3. Fragments (1-3 mm) of the alloy pieces were introduced into the apparatus, and the apparatus was sealed. The interior of the apparatus was adjusted to an atmosphere of I x 10-3 Pa, and the fragments were maintained in the atmosphere for three hours. Subsequently, the atmosphere was changed to a hydrogen atmosphere of 140 kPa, while the inside temperature was maintained 303 K. The change in pressure inside the apparatus was measured. Based on the obtained data, data of amount of hydrogen absorbed in the alloy at corresponding times were plotted on a graph, to thereby obtain time-elapsed change in amount of absorbed hydrogen. The period of time between the start of hydrogen pressurization and the point of time when the amount of absorbed hydrogen reached 1 % the maximum absorbable amount of hydrogen in the alloy (hereinafter the period is abbreviated as"T") and the maximum hydrogen-absorption rate (hereinafter abbreviated as"rmax") were calculated. The thus-calculated T and rmax were found to be 1,320 seconds and 4.6 x 104 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., residual magnetic flux density (hereinafter abbreviated as"Br"), coercive force (hereinafter abbreviated as"iHc"), and magnetic energy product (hereinafter abbreviated as"BHmax"), were investigated and found to be 1.37 T, 812 kA/m, and 375 kJ/m3, respectively.

(Example 2) In a manner similar to that employed in Example 1, a melt of the alloy A was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1, 000°C to be 800°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,580 seconds and 3.3 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 lem. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and. BHmax, were investigated and found to be 1.35 T, 788 kA/m, and 355 kJ/m3, respectively.

(Example 3) In a manner similar to that employed in Example 1, a melt of the alloy A was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container. Through cooling in the container, the average cooling rate within 800- 600°C was controlled to 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,090 seconds and 5.4 x 10 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 Hm. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.38 T, 828 kA/m, and 376 kJ/m3, respectively.

(Example 4) In a manner similar to that employed in Example 1, a melt of the alloy A was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,320 seconds and 4.0 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.36 T, 788 kA/m, and 360 kJ/m3, respectively.

(Comparative Example 1) In a manner similar to that employed in Example 1, a melt of the alloy A was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1, 000°C to 800°C/sec. The thus-cast alloy was collected in a product container, while helium gas was allowed to flow through the interior of the container so as to cool the alloy very rapidly. Through the gas flow, the average cooling rate within 800- 600°C was controlled to 15°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 2,540 seconds and 7.6 x 10-5 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 urn. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.32 T, 716 kA/m, and 347 kJ/m3, respectively.

(Comparative Example 2) In a manner similar to that employed in Example 1, a melt of the alloy A was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.7 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 300°C/sec. The thus-cast alloy was collected in a product container, while the container was maintained under reduced pressure so as to retard the cooling rate. Through the procedure, the average cooling rate within 800- 600°C was controlled to 0.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.40 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment identical with those of Example 1. The calculated T and rmax were found to be 170 seconds and 1. 9 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1. 30 T, 676 kA/m, and 337 kJ/m3, respectively.

(Comparative Example 3) In a manner similar to that employed in Example 1, a melt of the alloy A was prepared. The molten alloy was poured into a box-like mold (thickness: 20 mm) for casting (Book Mold method). The time required for cooling the molten alloy from the initial temperature to 1,000°C was measured, and the average cooling rate was calculated to be 8°C/sec. The average cooling rate within 800-600°C was controlled to 0. 1 °C/sec.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 60 seconds and 2.5 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.25 T, 629 kA/m, and 311 kJ/m3, respectively.

In connection with alloy A described in the aforementioned Examples 1 to 4 and Comparative Examples 1 to 3, Fig. 5A is a graph showing the relationship between T and BHmax, and Fig. 5B is a graph showing the relationship between rmax and BHmax.

Herein, T represents the period of time from the instant when an ingot of the rare earth magnet alloy was placed in a hydrogen atmosphere to the point of time when the amount of absorbed hydrogen reached 1 % the maximum absorbable amount of hydrogen in the alloy ingot, and rmax represents the maximum hydrogen-absorption rate of the alloy ingot.

In Figs. 5A and 5B, black dots represent the results obtained in Examples 1 to 4, and white squares represent those obtained in Comparative Examples 1 to 3.

Figs. 5A and 5B indicate that magnets produced from the alloy ingots which exhibit a T falling within a range of 100-1,800 seconds and an rmax falling within a range of 1.2 x 10- 3 to 1.5 x 10-2 mass%/sec provide a magnet characteristic more excellent than that of magnets produced from the alloy ingots falling outside the above ranges.

(Example 5) Alloy raw materials were provided and mixed to thereby obtain the following composition: Nd = 33.4 mass%; B = 1.1 mass%; Al = 0.4 mass%; Cu = 0.03 mass%; and iron = balance (hereinafter the alloy of this composition is referred to as alloy B). The raw material mixture was placed in an alumina crucible and heated in a vacuum chamber, initially in a vacuum atmosphere, then in an argon gas atmosphere, to thereby produce a molten alloy. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a copper roll for rapid cooling. The peripheral velocity of the roll was controlled to 1.2 m/sec. The average cooling rate was determined in a manner similar to that employed in Example 1. Through this procedure, the average cooling rate within a range of the temperature of the molten alloy to 1,000°C was found to be 800°C/sec. The thus-cast alloy was collected in a container for accommodating the alloy.

The average cooling rate was found to be 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The same apparatus and conditions for hydrogen decrepitation as those of Example 1 was employed. The calculated T and rmax were found to be 380 seconds and 6.7 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.27 T, 836 kA/m, and 321 kJ/m3, respectively.

(Example 6) In a manner similar to that employed in Example 5, a melt of the alloy B was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 800°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 570 seconds and 4.5 x 104 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 pm. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.25 T, 804 kA/m, and 311 kJ/m3, respectively.

(Example 7) In a manner similar to that employed in Example 5, a melt of the alloy B was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container. Through cooling in the container, the average cooling rate within 800- 600°C was controlled to 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 280 seconds and 8.3 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 urn. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.29 T, 867 kA/m, and 331 kJ/m3, respectively.

(Example 8) In a manner similar to that employed in Example 5, a melt of the alloy B was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1, 000°C to be 400°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 470 seconds and 5.6 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.26 T, 796 kA/m, and 316 kJ/m3, respectively.

(Comparative Example 4) In a manner similar to that employed in Example 5, a melt of the alloy B was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to 800°C/sec. The thus-cast alloy was collected in a product container, while helium gas was allowed to flow through the interior of the container so as to cool the alloy very rapidly. Through the gas flow, the average cooling rate within 800- 600°C was controlled to 15°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 960 seconds and 1.3 x 10-5 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.23 T, 788 kA/m, and 301 kJ/m3, respectively.

(Comparative Example 5) In a manner similar to that employed in Example 5, a melt of the alloy B was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.7 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1, 000°C to be 300°C/sec. The thus-cast alloy was collected in a product container, while the container was maintained under reduced pressure so as to retard the cooling rate. Through the procedure, the average cooling rate within 800- 600°C was controlled to 0.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.40 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment identical with those of Example 1. The calculated T and rmax were found to be 90 seconds and 2.3 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 u. m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.21 T, 796 kA/m, and 286 kJ/m3, respectively.

(Comparative Example 6) In a manner similar to that employed in Example 5, a melt of the alloy B was prepared. The molten alloy was poured into a box-like mold (thickness: 20 mm) for casting. The time required for cooling the molten alloy from the initial temperature to 1, 000°C was measured, and the average cooling rate was calculated to be 8°C/sec. The average cooling rate within 800-600°C was controlled to 0.1 °C/sec.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 40 seconds and 3.1 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 nm. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet ; i. e., Br, iHc, and BHmax, were investigated and found to be 1.20 T, 716 kA/m, and 286 kJ/m3, respectively.

In connection with alloy B described in the aforementioned Examples 5 to 8 and Comparative Examples 4 to 6, Fig. 6A is a graph showing the relationship between T and BHmax, and Fig. 6B is a graph showing the relationship between rmax and BHmax.

Herein, T represents the period of time from the instant when an ingot of the rare earth magnet alloy was placed in a hydrogen atmosphere to the point of time when the amount of absorbed hydrogen reached 1% the maximum absorbable amount of hydrogen in the alloy ingot, and rmax represents the maximum hydrogen-absorption rate of the alloy ingot.

In Figs. 6A and 6B, black dots represent the results obtained in Examples 5 to 8, and white squares represent those obtained in Comparative Examples 4 to 6.

Figs. 6A and 6B indicate that magnets produced from the alloy ingots which exhibit a T falling within a range of 100-1,800 seconds and an rmax falling within a range of 1. 2 x 10-3 to 1.5 x 10-2 mass%/sec provide a magnet characteristic more excellent than that of magnets produced from the alloy ingots falling outside the above ranges.

(Example 9) Alloy raw materials were provided and mixed to thereby obtain the following composition: Nd = 29.2 mass%; B = 0.97 mass%; Al = 0.4 mass%; Cu = 0.03 mass%; and iron = balance (hereinafter the alloy of this composition is referred to as alloy C). The raw material mixture was placed in an alumina crucible and heated in a vacuum chamber, initially in a vacuum atmosphere, then in an argon gas atmosphere, to thereby produce a molten alloy. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a copper roll for rapid cooling. The peripheral velocity of the roll was controlled to 1.2 m/sec. The average cooling rate was determined in a manner similar to that employed in Example 1. Through this procedure, the average cooling rate within a range of the temperature of the molten alloy to 1,000°C was found to be 800°C/sec. The thus-cast alloy was collected in a container for accommodating the alloy.

The average cooling rate was found to be 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The same apparatus and conditions for hydrogen decrepitation as those of Example 1 was employed. The calculated T and rmax were found to be 1,410 seconds and 3.8 x 104 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.38 T, 804 kA/m, and 379 kJ/m3, respectively.

(Example 10) In a manner similar to that employed in Example 9, a melt of the alloy C was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1. 2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 800°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,690 seconds and 2.2 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.35 T, 764 kA/m, and 363 kJ/m3, respectively.

(Example 11) In a manner similar to that employed in Example 9, a melt of the alloy C was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container. Through cooling in the container, the average cooling rate within 800- 600°C was controlled to 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,200 seconds and 4.7 x 104 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.39 T, 820 kA/m, and 384 kJ/m3, respectively.

(Example 12) In a manner similar to that employed in Example 9, a melt of the alloy C was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,550 seconds and 3.0 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.37 T, 772 kA/m, and 373 kJ/m3, respectively.

(Comparative Example 7) In a manner similar to that employed in Example 9, a melt of the alloy C was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to 800°C/sec. The thus-cast alloy was collected in a product container, while helium gas was allowed to flow through the interior of the container so as to cool the alloy very rapidly. Through the gas flow, the average cooling rate within 800- 600°C was controlled to 15°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 3,040 seconds and 8.8 x 10-5 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 u. m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.33 T, 621 kA/m, and 352 kJ/m3, respectively.

(Comparative Example 8) In a manner similar to that employed in Example 9, a melt of the alloy C was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.7 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 300°C/sec. The thus-cast alloy was collected in a product container, while the container was maintained under reduced pressure so as to retard the cooling rate. Through the procedure, the average cooling rate within 800- 600°C was controlled to 0.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.40 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment identical with those of Example 1. The calculated T and rmax were found to be 150 seconds and 1.6 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 pm. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.30 T, 637 kA/m, and 337 kJ/m3, respectively.

(Comparative Example 9) In a manner similar to that employed in Example 9, a melt of the alloy C was prepared. The molten alloy was poured into a box-like mold (thickness: 20 mm) for casting. The time required for cooling the molten alloy from the initial temperature to 1,000°C was measured, and the average cooling rate was calculated to be 8°C/sec. The average cooling rate within 800-600°C was controlled to 0.1 °C/sec.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 90 seconds and 2.2 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.24 T, 573 kA/m, and 306 kJ/m3, respectively.

In connection with alloy C described in the aforementioned Examples 9 to 12 and Comparative Examples 7 to 9, Fig. 7A is a graph showing the relationship between T and BHmax, and Fig. 7B is a graph showing the relationship between rmax and BHmax.

Herein, T represents the period of time from the instant when an ingot of the rare earth magnet alloy was placed in a hydrogen atmosphere to the point of time when the amount of absorbed hydrogen reached 1 % the maximum absorbable amount of hydrogen in the alloy ingot, and rmax represents the maximum hydrogen-absorption rate of the alloy ingot.

In Figs. 7A and 7B, black dots represent the results obtained in Examples 9 to 12, and white squares represent those obtained in Comparative Examples 7 to 9.

Figs. 7A and 7B indicate that magnets produced from the alloy ingots which exhibit a T falling within a range of 100-1,800 seconds and an rmax falling within a range of 1.2 x 10-3 to 1.5 x 10-2 mass%/sec provide a magnet characteristic more excellent than that of magnets produced from the alloy ingots falling outside the above ranges.

(Example 13) Alloy raw materials were provided and mixed to thereby obtain the following composition: Nd = 27.5 mass%; Dy = 2.5 mass%; B = 0.98 mass%; Al = 0.3 mass%; Cu = 0.03 mass%; and iron = balance (hereinafter the alloy of this composition is referred to as alloy D). The raw material mixture was placed in an alumina crucible and heated in a vacuum chamber, initially in a vacuum atmosphere, then in an argon gas atmosphere, to thereby produce a molten alloy. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a copper roll for rapid cooling. The peripheral velocity of the roll was controlled to 1.2 m/sec. The average cooling rate was determined in a manner similar to that employed in Example 1. Through this procedure, the average cooling rate within. a range of the temperature of the molten alloy to 1,000°C was found to be 800°C/sec. The thus-cast alloy was collected in a container for accommodating the alloy. The average cooling rate was found to be 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The same apparatus and conditions for hydrogen decrepitation as those of Example 1 was employed. The calculated T and rmax were found to be 1,610 seconds and 4.1 x 10'"' mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.32 T, 1,289 kA/m, and 328 kJ/m3, respectively.

(Example 14) In a manner similar to that employed in Example 13, a melt of the alloy D was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1. 2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 800°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,900 seconds and 2.8 x 104 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.30 T, 1,265 kA/m, and 318 kJ/m3, respectively.

(Example 15) In a manner similar to that employed in Example 13, a melt of the alloy D was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0. 8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container. Through cooling in the container, the average cooling rate within 800- 600°C was controlled to 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,390 seconds and 4.9 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.33 T, 1,305 kA/m, and 333 kJ/m3, respectively.

(Example 16) In a manner similar to that employed in Example 13, a melt of the alloy D was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,630 seconds and 3.5 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.31 T, 1,273 kA/m, and 323 kJ/m3, respectively.

(Comparative Example 10) In a manner similar to that employed in Example 13, a melt of the alloy D was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to 800°C/sec. The thus-cast alloy was collected in a product container, while helium gas was allowed to flow through the interior of the container so as to cool the alloy very rapidly. Through the gas flow, the average cooling rate within 800- 600°C was controlled to 15°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 3,030 seconds and 6.4 x 10-5 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 pm. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.27 T, 1,218 kA/m, and 304 kJ/m3, respectively.

(Comparative Example 11) In a manner similar to that employed in Example 13, a melt of the alloy D was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.7 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1, 000°C to be 300°C/sec. The thus-cast alloy was collected in a product container, while the container was maintained under reduced pressure so as to retard the cooling rate. Through the procedure, the average cooling rate within 800- 600°C was controlled to 0.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.40 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment identical with those of Example 1. The calculated T and rmax were found to be 180 seconds and 1.4 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.25 T, 1,202 kA/m, and 295 kJ/m3, respectively.

(Comparative Example 12) In a manner similar to that employed in Example 13, a melt of the alloy D was prepared. The molten alloy was poured into a box-like mold (thickness: 20 mm) for casting. The time required for cooling the molten alloy from the initial temperature to 1, 000°C was measured, and the average cooling rate was calculated to be 8°C/sec. The average cooling rate within 800-600°C was controlled to 0.1 °C/sec.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 80 seconds and 2.1 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.20 T, 1,162 kA/m, and 273 kJ/m3, respectively.

In connection with alloy D described in the aforementioned Examples 13 to 16 and Comparative Examples 10 to 12, Fig. 8A is a graph showing the relationship between T and BHmax, and Fig. 8B is a graph showing the relationship between rmax and BHmax.

Herein, T represents the period of time from the instant when an ingot of the rare earth magnet alloy was placed in a hydrogen atmosphere to the point of time when the amount of absorbed hydrogen reached 1 % the maximum absorbable amount of hydrogen in the alloy ingot, and rmax represents the maximum hydrogen-absorption rate of the alloy ingot.

In Figs. 8A and 8B, black dots represent the results obtained in Examples 13 to 16, and white squares represent those obtained in Comparative Examples 10 to 12.

Figs. 8A and 8B indicate that magnets produced from the alloy ingots which exhibit a T falling within a range of 200-2,400 seconds and an rmax falling within a range of 1.0 x 10-3 to 1.2 x 10'2 mass%/sec provide a magnet characteristic more excellent than that of magnets produced from the alloy ingots falling outside the above ranges.

(Example 17) Alloy raw materials were provided and mixed to thereby obtain the following composition: Nd = 31. 9 mass%; Dy = 1.5 mass%; B = 1.1 mass%; Al = 0.4 mass%; Cu = 0.03 mass%; and iron = balance (hereinafter the alloy of this composition is referred to as alloy E). The raw material mixture was placed in an alumina crucible and heated in a vacuum chamber, initially in a vacuum atmosphere, then in an argon gas atmosphere, to thereby produce a molten alloy. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a copper roll for rapid cooling. The peripheral velocity of the roll was controlled to 1.2 m/sec. The average cooling rate was determined in a manner similar to that employed in Example 1. Through this procedure, the average cooling rate within a range of the temperature of the molten alloy to 1,000°C was found to be 800°C/sec. The thus-cast alloy was collected in a container for accommodating the alloy. The average cooling rate was found to be 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The same apparatus and conditions for hydrogen decrepitation as those of Example 1 was employed. The calculated T and rmax were found to be 700 seconds and 6.2 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.25 T, 1,074 kA/m, and 292 kJ/m3, respectively.

(Example 18) In a manner similar to that employed in Example 17, a melt of the alloy E was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 800°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 880 seconds and 4.2 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3. 2 llm. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.22 T, 1,058 kA/m, and 279 kJ/m3, respectively.

(Example 19) In a manner similar to that employed in Example 17, a melt of the alloy E was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container. Through cooling in the container, the average cooling rate within 800- 600°C was controlled to 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 590 seconds and 8.0 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.27 T, 1,114 kA/m, and 302 kJ/m3, respectively.

(Example 20) In a manner similar to that employed in Example 17, a melt of the alloy E was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 810 seconds and 5.3 x 10'4 mass%/sec, respectively.

$131-1 The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.23 T, 1,074 kA/m, and 283 kJ/m3, respectively.

(Comparative Example 13) In a manner similar to that employed in Example 17, a melt of the alloy E was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to 800°C/sec. The thus-cast alloy was collected in a product container, while helium gas was allowed to flow through the interior of the container so as to cool the alloy very rapidly. Through the gas flow, the average cooling rate within 800- 600°C was controlled to 15°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,430 seconds and 1. 1 x 10-5 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.20 T, 1,035 kA/m, and 270 kJ/m3, respectively.

(Comparative Example 14) In a manner similar to that employed in Example 17, a melt of the alloy E was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.7 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 300°C/sec. The thus-cast alloy was collected in a product container, while the container was maintained under reduced pressure so as to retard the cooling rate. Through the procedure, the average cooling rate within 800- 600°C was controlled to 0.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.40 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment identical with those of Example 1. The calculated T and rmax were found to be 150 seconds and 2.0 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.18 T, 1,042 kA/m, and 261 kJ/m3, respectively.

(Comparative Example 15) In a manner similar to that employed in Example 17, a melt of the alloy E was prepared. The molten alloy was poured into a box-like mold (thickness: 20 mm) for casting. The time required for cooling the molten alloy from the initial temperature to 1,000°C was measured, and the average cooling rate was calculated to be 8°C/sec. The average cooling rate within 800-600°C was controlled to 0.1 °C/sec.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 50 seconds and 2.9 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.16 T, 955 kA/m, and 252 kJ/m3, respectively.

In connection with alloy E described in the aforementioned Examples 17 to 20 and Comparative Examples 13 to 15, Fig. 9A is a graph showing the relationship between T and BHmax, and Fig. 9B is a graph showing the relationship between rmax and BHmax.

Herein, T represents the period of time from the instant when an ingot of the rare earth magnet alloy was placed in a hydrogen atmosphere to the point of time when the amount of absorbed hydrogen reached 1 % the maximum absorbable amount of hydrogen in the alloy ingot, and rmax represents the maximum hydrogen-absorption rate of the alloy ingot.

In Figs. 9A and 9B, black dots represent the results obtained in Examples 17 to 20, and white squares represent those obtained in Comparative Examples 13 to 15.

$138-1 Figs. 9A and 9B indicate that magnets produced from the alloy ingots which exhibit a T falling within a range of 200-2,400 seconds and an rmax falling within a range of 1.0 x 10-3 to 1.2 x 10-2 mass%/sec provide a magnet characteristic more excellent than that of magnets produced from the alloy ingots falling outside the above ranges.

(Example 21) Alloy raw materials were provided and mixed to thereby obtain the following composition: Nd = 25.2 mass%; Dy = 4.0 mass%; B = 0.97 mass%; Al = 0.3 mass%; Cu = 0.03 mass%; and iron = balance (hereinafter the alloy of this composition is referred to as alloy F). The raw material mixture was placed in an alumina crucible and heated in a vacuum chamber, initially in a vacuum atmosphere, then in an argon gas atmosphere, to thereby produce a molten alloy. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a copper roll for rapid cooling. The peripheral velocity of the roll was controlled to 1.2 m/sec. The average cooling rate was determined in a manner similar to that employed in Example 1. Through this procedure, the average cooling rate within a range of the temperature of the molten alloy to 1,000°C was found to be 800°C/sec. The thus-cast alloy was collected in a container for accommodating the alloy. The average cooling rate was found to be 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The same apparatus and conditions for hydrogen decrepitation as those of Example 1 was employed. The calculated T and rmax were found to be 1,750 seconds and 3.2 x 104 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.30 T, 1,560 kA/m, and 325 kJ/m3, respectively.

(Example 22) In a manner similar to that employed in Example 21, a melt of the alloy F was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 800°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,990 seconds and 1.7 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 u. m. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.27 T, 1,520 kA/m, and 305 kJ/m3, respectively.

(Example 23) In a manner similar to that employed in Example 21, a melt of the alloy F was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container. Through cooling in the container, the average cooling rate within 800- 600°C was controlled to 0.5°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,550 seconds and 4.1 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.31 T, 1,576 kA/m, and 325 kJ/m3, respectively.

(Example 24) In a manner similar to that employed in Example 21, a melt of the alloy F was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.8 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 400°C/sec. The thus-cast alloy was collected in a product container, while argon gas was allowed to flow through the interior of the container so as to cool the alloy more rapidly. Through the gas flow, the average cooling rate within 800-600°C was controlled to 1.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.35 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 1,780 seconds and 2.8 x 10-4 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 urn. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.28 T, 1,528 kA/m, and 310 kJ/m3, respectively.

(Comparative Example 16) In a manner similar to that employed in Example 21, a melt of the alloy F was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 1.2 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to 800°C/sec. The thus-cast alloy was collected in a product container, while helium gas was allowed to flow through the interior of the container so as to cool the alloy very rapidly. Through the gas flow, the average cooling rate within 800- 600°C was controlled to 15°C/sec. The obtained cast alloy pieces had an average thickness of 0.23 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 3,360 seconds and 7.6 x 10-5 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 Fm. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.25 T, 1,393 kA/m, and 294 kJ/m3, respectively.

(Comparative Example 17) In a manner similar to that employed in Example 21, a melt of the alloy F was prepared and cast. Upon casting, the molten alloy was poured into a tundish, and then poured from the tundish onto a roll. The peripheral velocity of the roll was controlled to 0.7 m/sec, to thereby regulate the average cooling rate within a range of the temperature of the molten alloy to 1,000°C to be 300°C/sec. The thus-cast alloy was collected in a product container, while the container was maintained under reduced pressure so as to retard the cooling rate. Through the procedure, the average cooling rate within 800- 600°C was controlled to 0.2°C/sec. The obtained cast alloy pieces had an average thickness of 0.40 mm.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment identical with those of Example 1. The calculated T and rmax were found to be 180 seconds and 1.3 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3.2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.22 T, 1,377 kA/m, and 280 kJ/m3, respectively.

(Comparative Example 18) In a manner similar to that employed in Example 21, a melt of the alloy F was prepared. The molten alloy was poured into a box-like mold (thickness: 20 mm) for casting. The time required for cooling the molten alloy from the initial temperature to 1,000°C was measured, and the average cooling rate was calculated to be 8°C/sec. The average cooling rate within 800-600°C was controlled to 0.1 °C/sec.

Subsequently, the thus-obtained cast alloy was subjected to hydrogen decrepitation.

The apparatus and conditions employed for carrying out the treatment were identical with those of Example 1. The calculated T and rmax were found to be 120 seconds and 1.8 x 10-3 mass%/sec, respectively.

The thus-treated alloy was further pulverized into a powder having an average particle size, as measured by use of Fisher Sub-Sieve Sizer, of 3. 2 um. The powder was molded in a magnetic field, and the resultant compact was sintered, to thereby produce a sintered magnet. Magnetic characteristics of the magnet; i. e., Br, iHc, and BHmax, were investigated and found to be 1.16 T, 1,337 kA/m, and 252 kJ/m3, respectively.

In connection with alloy F described in the aforementioned Examples 21 to 24 and Comparative Examples 16 to 18, Fig. I OA is a graph showing the relationship between T and BHmax, and Fig. 10B is a graph showing the relationship between rmax and BHmax.

Herein, T represents the period of time from the instant when an ingot of the rare earth magnet alloy was placed in a hydrogen atmosphere to the point of time when the amount of absorbed hydrogen reached 1% the maximum absorbable amount of hydrogen in the alloy ingot, and rmax represents the maximum hydrogen-absorption rate of the alloy ingot.

In Figs. I OA and 10B, black dots represent the results obtained in Examples 21 to 24, and white squares represent those obtained in Comparative Examples 16 to 18.

Figs. I OA and 10B indicate that magnets produced from the alloy ingots which exhibit a T falling within a range of 200-2,400 seconds and an rmax falling within a range of 1.0 x 10-3 to 1.2 x 10'2 mass%/sec provide a magnet characteristic more excellent than that of magnets produced from the alloy ingots falling outside the above ranges.

Table 1 shows the compositions of the rare earth magnet alloys which have been used in the aforementioned Examples 1 to 24 and Comparative Examples 1 to 18.

Table 1 Nd Dy B Al Cu Fe alloy A 30. 0-0. 98 0. 3 0. 03 68. 69 alloy B 33. 4 1. 10 0. 4 0. 03 65. 07 alloy C 29. 2-0. 97 0. 3 0. 03 69. 5 alloy D 27. 5 2. 5 0. 98 0. 3 0. 03 68. 69 alloy E 31.9 1.5 1.10 0.4 0.03 65. 07 alloy F 25. 2 4. 0 0. 97 0. 3 0. 03 69. 5 Tables 2-1 to 2-3 show characteristics of alloys A, B, and C investigated in Examples 1 to 12 and Comparative Examples 1 to 9.

Table 2-1 Peripheral Alloy Casting Cooling rate Average rmax Velocity T Br iHc BHmax composition method °C/sec thickness of roll mass% 1000- 800- sec T kA/m kJ/m3 m/s mm /sec mp 600 Ex. 1 A SC 1.2 800 0.5 0.23 1320 4.60E-04 1.37 811.69 374.81 Ex. 2 A SC 1.2 800 1.2 0.23 1580 3.30E-04 1.35 787.82 354.92 Ex. 3 A SC 0.8 400 0.5 0.35 1090 5.40E-04 1.38 827.61 376.40 Ex. 4 A SC 0.8 400 1.2 0.35 1320 4.00E-04 1.36 787.82 359.69 Comp. A SC 1.2 800 15 0.23 2540 7.60E-05 1.32 716.20 346.96 Ex. 1 Comp. A SC 0.7 300 0.2 0.40 170 1.90E-03 1.30 676.41 336.61 Ex. 2 Comp. A BM - - - - 60 2.50E-03 1.25 628.66 311.15 Ex. 3 Table 2-2 Peripheral Alloy Casting Cooling rate Av. max velocity T Br iHc BHmax composition method °C/sec thickness of roll mass% 800- T kA/m kJ/m3 m/s 100-mp mm /sec 600 Ex. 5 B SC 1.2 800 0.5 0.23 380 6.70E-04 1.27 835.56 320.70 Ex. 6 B SC 1.2 800 1.2 0.23 570 4.50E-04 1.25 803.73 311.15 Ex. 7 B SC 0.8 400 0.5 0.35 280 8.30E-04 1.29 867.39 331.04 Ex. 8 B SC 0.8 400 1.2 0.35 470 5.60E-04 1.26 795.77 315.92 Comp. B SC 1.2 800 15 0.23 960 1.30E-05 1.23 787.82 300.80 Ex. 4 Comp. B SC 0.7 300 0.2 0.40 90 2.30E-03 1.21 795.77 291.25 Ex. 5 Comp. B BM - - - - 40 3.10E-03 1.20 716.20 286.48 Ex. 6 Table 2-3 Peripheral Alloy Casting Cooling rate Av. rmax velocity T Br iHc BHmax composition method °C/sec thickness of roll mass% 1000- 800- sec T kA/m kJ/m3 m/s mm /sec mp 600 Ex. 9 C SC 1.2 800 0.5 0.23 1410 3.80E-04 1.38 803.73 378.79 Ex. 10 C SC 1.2 800 1.2 0.23 1690 2.20E-04 1.25 763.94 362.87 EX. 11 C SC 0.8 400 0.5 0.35 1200 4.70E-04 1.39 819.65 384.36 Ex. 12 C SC 0.8 400 1.2 0.35 1550 3.00E-04 1.37 771.90 373.22 Comp. C SC 1.2 800 15 0.23 3040 8.80e-05 1.33 620.70 351.73 Ex. 7 Comp. C SC 0.7 300 0.2 0.40 150 1.60E-03 1.30 636.62 336.61 Ex. 8 Comp. C BM - - - - 90 2.20E-03 1.24 572.96 305.58 Ex. 9 Tables 3-1 to 3-3 show characteristics of alloys D, E, and F investigated in Examples 13 to 24 and Comparative Examples 10 to 18.<BR> <P>Table 3-1 Peripheral Alloy Casting Cooling rate Av. rmax velocity T Br iHc BHmax composition method °C/sec thickness of roll mass% 1000- 800- sec T kA/m kJ/m3 m/s mm /sec mp 600 Ex. 13 D SC 1.2 800 0.5 0.23 1610 4.10E-04 1.32 1,289.16 327.86 Ex. 14 D SC 1.2 800 1.2 0.23 1900 2.80E-04 1.30 1,265.28 318.31 Ex. 15 D SC 0.8 400 0.5 0.35 1390 4.90E-04 1.33 1,305.07 332.63 Ex. 16 D SC 0.8 400 1.2 0.35 1630 3.50E-04 1.31 1,273.24 323.08 Comp. D SC 1.2 800 15 0.23 3030 6.40E-05 1.27 1,217.54 303.99 Ex. 10 Comp. D SC 0.7 300 0.2 0.40 180 1.40E-03 1.25 1,201.62 295.23 Ex. 11 Comp. D BM - - - - 80 2.10E-03 1.20 1,161.83 272.95 Ex. 12 Table 3-2 Peripheral Alloy Casting Cooling rate Av. max velocity T Br iHc BHmax composition method °C/sec thickness of roll mass% 1000- 800- sec T kA/m kJ/m3 m/s mm /sec mp 600 Ex. 17 E SC 1.2 800 0.5 0.23 700 6.20E-04 1.25 1,074.30 292.05 Ex. 18 E SC 1.2 800 1.2 0.23 880 4.20E-04 1.22 1,058.38 278.52 Ex. 19 E SC 0.8 400 0.5 0.35 590 8.00E-04 1.27 1,114.08 301.60 Ex. 20 E SC 0.8 400 1.2 0.35 810 5.30E-04 1.23 1,074.30 283.30 Comp. E SC 1.2 800 15 0.23 1430 1.10E-05 1.20 1,034.51 269.77 Ex. 13 Comp. E SC 0.7 300 0.2 0.40 150 2.00E-03 1.18 1,042.46 261.01 Ex. 14 Comp. E BM - - - - 50 2.90E-03 1.16 954.93 252.26 Ex. 15 Table 3-3 Peripheral Alloy Casting Cooling rate Av. rmax velocity T Br iHc BHmax compositon method °C/sec thickness of roll mass% 1000- 800- sec T kA/m kJ/m3 m/s mm /sec mp 600 Ex. 21 F SC 1.2 800 0.5 0.23 1750 3.20E-04 1.30 1,559.72 324.68 Ex. 22 F SC 1.2 800 1.2 0.23 1990 1.70E-04 1.27 1,519.93 304.78 Ex. 23 F SC 0.8 400 0.5 0.35 1550 4.10E-04 1.31 1,575.63 324.68 Ex. 24 F SC 0.8 400 1.2 0.35 1780 2.80E-04 1.28 1,527.89 309.56 Comp. F SC 1.2 800 15 0.23 3360 7.60E-05 1.25 1,392.61 294.44 Ex. 16 Comp. F SC 0.7 300 0.2 0.40 180 1.30E-03 1.22 1.376.69 280.11 Ex. 17 Comp. F BM - - - - 120 1.80E-03 1.16 1,336.90 252.26 Ex. 18 INDUSTRIAL APPLICABILITY In contrast to a conventional method for evaluating metallographic structure of an alloy only on the basis of a cross-section photograph; e. g., determination of the inter R- rich phase spacing by use of a microscopic photograph of a rare earth magnet alloy ingot and image processing of the photograph, the present invention, employing a novel method including evaluation of metallographic structure of a rare earth magnet alloy ingot on the basis of hydrogen absorption characteristics, can be applied to quality determination of large amounts of alloy ingots as compared with a conventional method, and can evaluate not just a portion of the alloy but the entirety of the alloy. Thus, the particle size distribution of the alloy powder after completion of hydrogen decrepitation and magnetic characteristics of the sintered magnet produced from the powder can be accurately predicted by evaluating the rare earth magnet alloy ingot itself, and quality of the ingot can be determined.

The method of the present invention for determining quality of a rare earth magnet alloy ingot identifies the conditions of the hydrogen absorption behavior of a rare earth magnet alloy ingot from which an alloy powder suitable for fabricating a magnet of more excellent characteristics can be produced, and can predict the quality of a rare earth magnet alloy powder and magnetic characteristics by evaluating the rare earth magnet alloy ingot itself, more accurately as compared with a conventional method including evaluating the distribution of RE-rich phase on the basis of a cross-section photograph.

Thus, the quality of a rare earth magnet alloy ingot can be determined by evaluating the ingot itself. Conventionally, the quality of a rare earth magnet alloy ingot cannot be evaluated until the final sintered magnet has been produced. However, according to the present invention, the quality of the ingot can be determined by evaluating the ingot itself, to thereby shorten the time required for the production step of a rare earth magnet alloy ingot, leading to reduction of costs.

In addition, a rare earth magnet produced from a rare earth magnet alloy which has been qualified on the basis of the method of the present invention for determining quality of a rare earth magnet alloy ingot exhibits excellent magnetic characteristics.